Patients undergo video-EEG monitoring to determine seizure type and focus localization. Positron emission tomography (PET) and magnetic resonance imaging (MRI) are used to study cerebral metabolism, blood flow, and structure. We compared-regional brain volumes in patients with temporal lobe epilepsy to the same measures in controls. Hippocampal volumes ipsilateral to the epileptogenic zone were significantly smaller in the patients with epilepsy as compared to controls. In addition, mean ipsilateral thalamic, caudate and bilateral lenticular volumes were significantly smaller in the patients with epilepsy as compared to controls. The ipsilateral/contralateral thalamic volume ratio was also significantly lower in the patients with epilepsy as compared to controls (p<0.0001), but there were no significant group differences in caudate or lenticular ratios. These results show that medically intractable temporal lobe epilepsy is associated with volume loss in brain structures outside the presumably involved hippocampus. The pathophysiologic significance of these findings are unclear. They could be related to the underlying etiology of the disorder. However, the volume loss may reflect damage due to the involvement of these structures in recurrent seizure activity. We also found that epilepsy duration had a weak but significant relation to ipsilateral hippocampal formation (HF) volume and ipsilateral/contralateral ratio. Patients with a history of prolonged complex febrile seizures (FS) had smaller ipsilateral HF volume and ipsilateral/contralateral ratio than patients without a history of FS. The effect of age at onset or scan was not significant. Patients with a history of FS did not have earlier age at epilepsy onset or longer duration than those without a history of FS. A history of FS predicted the severity of HF atrophy in our patients. Age at onset or study was not a significant factor. Epilepsy duration, however, did have a weak but significant effect, suggesting that, after an initial insult, progressive HF damage may occur in patients with persistent seizures. We are conducting a double-blind placebo controlled study of the effect of low-frequency (1/second) transcranial magnetic stimulation (TMS) on the frequency of complex partial seizures. Animal studies suggest that low frequency TMS may induce long-term depression of neuronal excitability. Patients whose seizures are not controlled by currently available anti- epileptic drugs, and have a seizure focus localized by video-EEG monitoring are being enrolled. In the placebo condition, the magnetic coil is directed away from the seizure focus.